Run LGG and HGG survival by splicing cluster assignment and splicing burden

Purpose:

Runs survival analysis models using splicing cluster assignment and 1) single exon splicing burden index (SBI) or 2) KEGG Spliceosome GSVA scores as a predictor

Usage

Uses a wrapper function (survival_analysis) from utils folder.

Setup

Packages and functions

Load packages, set directory paths and call setup script

library(tidyverse)
library(survival)
library(ggpubr)
library(ggplot2)
library(patchwork)

root_dir <- rprojroot::find_root(rprojroot::has_dir(".git"))

data_dir <- file.path(root_dir, "data")
analysis_dir <- file.path(root_dir, "analyses", "survival")
input_dir <- file.path(analysis_dir, "results")
results_dir <- file.path(analysis_dir, "results")
plot_dir <- file.path(analysis_dir, "plots")

# If the input and results directories do not exist, create it
if (!dir.exists(results_dir)) {
  dir.create(results_dir, recursive = TRUE)
}

source(file.path(analysis_dir, "util", "survival_models.R"))

Set metadata and cluster assignment file paths

metadata_file <- file.path(input_dir, "splicing_indices_with_survival.tsv")

cluster_file <- file.path(root_dir, "analyses",
                          "sample-psi-clustering", "results",
                          "sample-cluster-metadata-top-5000-events-stranded.tsv")

kegg_scores_stranded_file <- file.path(root_dir, "analyses",
                          "sample-psi-clustering", "results",
                          "gsva_output_stranded.tsv")

Wrangle data Add cluster assignment and spliceosome gsva scores to metadata and define column lgg_group (LGG or non_LGG)

metadata <- read_tsv(metadata_file)

Rows: 684 Columns: 26
── Column specification ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Delimiter: "\t"
chr (11): Kids_First_Biospecimen_ID, Histology, Kids_First_Participant_ID, molecular_subtype, extent_of_tumor_resection, EFS_event_type, OS_...
dbl (15): Total, AS_neg, AS_pos, AS_total, SI_A3SS, SI_A5SS, SI_RI, SI_SE, SI_Total, EFS_days, OS_days, age_at_diagnosis_days, age_at_diagno...

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

clusters <- read_tsv(cluster_file) %>%
  dplyr::rename(Kids_First_Biospecimen_ID = sample_id)

Rows: 729 Columns: 8
── Column specification ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Delimiter: "\t"
chr (6): sample_id, plot_group, plot_group_hex, RNA_library, molecular_subtype, plot_group_n
dbl (2): cluster, group_n

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

gsva_scores <- read_tsv(kegg_scores_stranded_file) %>%
  dplyr::filter(geneset == "KEGG_SPLICEOSOME") %>%
  dplyr::rename(spliceosome_gsva_score = score)

Rows: 22599 Columns: 3
── Column specification ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Delimiter: "\t"
chr (2): sample_id, geneset
dbl (1): score

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

# how many clusters?
n_clust <- length(unique(clusters$cluster))

metadata <- metadata %>%
  right_join(clusters %>% dplyr::select(Kids_First_Biospecimen_ID,
                                       cluster)) %>%
  left_join(gsva_scores %>% dplyr::select(sample_id,
                                          spliceosome_gsva_score),
            by = c("Kids_First_Biospecimen_ID" = "sample_id")) %>% 
  dplyr::mutate(cluster = glue::glue("Cluster {cluster}")) %>%
  dplyr::mutate(cluster = fct_relevel(cluster,
                                               paste0("Cluster ", 1:n_clust))) %>%
  dplyr::mutate(lgg_group = case_when(
    plot_group == "Low-grade glioma" ~ "LGG",
    TRUE ~ "non-LGG"
  )) %>%
  dplyr::mutate(SBI = SI_Total * 10) %>%
  dplyr::mutate(age_at_diagnosis_years = age_at_diagnosis_days/365.25)

Joining with `by = join_by(Kids_First_Biospecimen_ID)`

Generate coxph models including extent of tumor resection, lgg group, and cluster assignment and SBI as covariates

add_model_os <- fit_save_model(metadata[!metadata$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                              terms = "extent_of_tumor_resection+lgg_group+cluster+age_at_diagnosis_years+SBI",
                               file.path(results_dir, "cox_OS_additive_terms_resection_lgg_group_cluster_SBI.RDS"),
                               "multivariate",
                               years_col = "OS_years",
                               status_col = "OS_status")

forest_os <- plotForest(readRDS(file.path(results_dir, "cox_OS_additive_terms_resection_lgg_group_cluster_SBI.RDS")))

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

forest_os

ggsave(file.path(plot_dir, "forest_add_OS_resection_lgg_group_cluster_assignment_SBI.pdf"),
       forest_os,
       width = 10, height = 6, units = "in",
       device = "pdf")

add_model_efs <- fit_save_model(metadata[!metadata$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                              terms = "extent_of_tumor_resection+lgg_group+cluster+age_at_diagnosis_years+SBI",
                               file.path(results_dir, "cox_EFS_additive_terms_resection_lgg_group_cluster_SBI.RDS"),
                               "multivariate",
                               years_col = "EFS_years",
                               status_col = "EFS_status")

forest_efs <- plotForest(readRDS(file.path(results_dir, "cox_EFS_additive_terms_resection_lgg_group_cluster_SBI.RDS")))

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

forest_efs

ggsave(file.path(plot_dir, "forest_add_EFS_resection_lgg_group_cluster_assignment_SBI.pdf"),
       forest_efs,
       width = 10, height = 6, units = "in",
       device = "pdf")

repeat analysis, replacing SBI with KEGG spliceosome gsva score

add_model_os <- fit_save_model(metadata[!metadata$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                              terms = "extent_of_tumor_resection+lgg_group+cluster+age_at_diagnosis_years+spliceosome_gsva_score",
                               file.path(results_dir, "cox_OS_additive_terms_resection_lgg_group_cluster_spliceosome_score.RDS"),
                               "multivariate",
                               years_col = "OS_years",
                               status_col = "OS_status")

forest_os <- plotForest(readRDS(file.path(results_dir, "cox_OS_additive_terms_resection_lgg_group_cluster_spliceosome_score.RDS")))

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

forest_os

ggsave(file.path(plot_dir, "forest_add_OS_resection_lgg_group_cluster_assignment_spliceosome_score.pdf"),
       forest_os,
       width = 10, height = 6, units = "in",
       device = "pdf")

models <- c("spliceosome_gsva_score", "SBI")

for (each in models) {
  int_model_efs <- fit_save_model(metadata[!metadata$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                                terms = paste0("extent_of_tumor_resection+lgg_group+cluster*", each, "+age_at_diagnosis_years"),
                                 file.path(results_dir, paste0("cox_EFS_interaction_terms_resection_lgg_group_cluster_", each, ".RDS")),
                                 "multivariate",
                                 years_col = "EFS_years",
                                 status_col = "EFS_status")
  
  int_forest_efs <- plotForest(readRDS(file.path(results_dir, paste0("cox_EFS_interaction_terms_resection_lgg_group_cluster_", each, ".RDS"))))
  
  int_forest_efs
  
  ggsave(file.path(plot_dir, paste0("forest_int_EFS_resection_lgg_group_cluster_assignment_", each, ".pdf")),
         int_forest_efs,
         width = 10, height = 6, units = "in",
         device = "pdf")

  int_model_os <- fit_save_model(metadata[!metadata$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                                terms = paste0("extent_of_tumor_resection+lgg_group+cluster*", each, "+age_at_diagnosis_years"),
                                 file.path(results_dir, paste0("cox_OS_interaction_terms_resection_lgg_group_cluster_", each, ".RDS")),
                                 "multivariate",
                                 years_col = "OS_years",
                                 status_col = "OS_status")
  
  int_forest_os <- plotForest(readRDS(file.path(results_dir, paste0("cox_OS_interaction_terms_resection_lgg_group_cluster_", each, ".RDS"))))
  
  int_forest_os
  
  ggsave(file.path(plot_dir, paste0("forest_int_OS_resection_lgg_group_cluster_assignment_", each, ".pdf")),
         int_forest_os,
         width = 10, height = 6, units = "in",
         device = "pdf")
  
}

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

add_model_efs <- fit_save_model(metadata[!metadata$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                              terms = "extent_of_tumor_resection+lgg_group+cluster+age_at_diagnosis_years+spliceosome_gsva_score",
                               file.path(results_dir, "cox_EFS_additive_terms_resection_lgg_group_cluster_spliceosome_score.RDS"),
                               "multivariate",
                               years_col = "EFS_years",
                               status_col = "EFS_status")

forest_efs <- plotForest(readRDS(file.path(results_dir, "cox_EFS_additive_terms_resection_lgg_group_cluster_spliceosome_score.RDS")))

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

forest_efs

ggsave(file.path(plot_dir, "forest_add_EFS_resection_lgg_group_cluster_assignment_spliceosome_score.pdf"),
       forest_efs,
       width = 10, height = 6, units = "in",
       device = "pdf")

Subset metadata for LGG, and only include clusters with >= 10 samples

lgg <- metadata %>%
  dplyr::filter(plot_group == "Low-grade glioma") %>%
  dplyr::mutate(cluster = factor(cluster)) %>%
  dplyr::mutate(mol_sub_group = fct_relevel(mol_sub_group, "Wildtype", after = 0))

retain_clusters_lgg <- lgg %>%
  dplyr::count(cluster) %>%
  filter(n >= 10) %>%
  pull(cluster)

lgg <- lgg %>%
  filter(cluster %in% retain_clusters_lgg) %>%
    dplyr::mutate(cluster = factor(cluster))

Generate coxph models including covariates extent_of_tumor_resection, mol_sub_group, cluster, and SBI and plot

# identify LGG clusters
lgg_clusters <- metadata %>%
  filter(lgg_group == "LGG") %>%
  mutate(cluster = as.integer(gsub("cluster", "", cluster))) %>%
  pull(cluster) %>%
  sort() %>%
  unique()

Warning: There was 1 warning in `mutate()`.
ℹ In argument: `cluster = as.integer(gsub("cluster", "", cluster))`.
Caused by warning:
! NAs introduced by coercion

add_model_lgg_efs <- fit_save_model(lgg[!lgg$cluster %in% lgg_clusters & !lgg$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                              terms = "extent_of_tumor_resection+mol_sub_group+cluster+age_at_diagnosis_years+SBI",
                               file.path(results_dir, "cox_lgg_EFS_additive_terms_resection_subtype_cluster_SBI.RDS"),
                               "multivariate",
                               years_col = "EFS_years",
                               status_col = "EFS_status")

Warning in coxph.fit(X, Y, istrat, offset, init, control, weights = weights,  :
  Loglik converged before variable  6,7 ; coefficient may be infinite. 

forest_lgg_efs <- plotForest(readRDS(file.path(results_dir, "cox_lgg_EFS_additive_terms_resection_subtype_cluster_SBI.RDS")))

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

forest_lgg_efs

ggsave(file.path(plot_dir, "forest_add_EFS_LGG_resection_subtype_cluster_assignment_SBI.pdf"),
       forest_lgg_efs,
       width = 10, height = 6, units = "in",
       device = "pdf")

repeat analysis replacing SBI with spliceosome_gsva_score

add_model_lgg_efs <- fit_save_model(lgg[!lgg$cluster %in% lgg_clusters & !lgg$extent_of_tumor_resection %in% c("Not Reported", "Unavailable"),],
                              terms = "extent_of_tumor_resection+mol_sub_group+cluster+age_at_diagnosis_years+spliceosome_gsva_score",
                               file.path(results_dir, "cox_lgg_EFS_additive_terms_resection_subtype_cluster_spliceosome_score.RDS"),
                               "multivariate",
                               years_col = "EFS_years",
                               status_col = "EFS_status")

Warning in coxph.fit(X, Y, istrat, offset, init, control, weights = weights,  :
  Loglik converged before variable  6,7 ; coefficient may be infinite. 

forest_lgg_efs <- plotForest(readRDS(file.path(results_dir, "cox_lgg_EFS_additive_terms_resection_subtype_cluster_spliceosome_score.RDS")))

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

forest_lgg_efs

ggsave(file.path(plot_dir, "forest_add_EFS_LGG_resection_subtype_cluster_assignment_spliceosome_score.pdf"),
       forest_lgg_efs,
       width = 10, height = 6, units = "in",
       device = "pdf")

Subset metadata for HGG and retain cluster with n >= 10

hgg <- metadata %>%
  dplyr::filter(plot_group %in% c("Other high-grade glioma", "Diffuse midline glioma")) %>%
  dplyr::mutate(cluster = factor(cluster)) %>%
  dplyr::mutate(mol_sub_group = fct_relevel(mol_sub_group, "HGG, H3 wildtype", after = 0)) %>%
  dplyr::filter(!is.na(OS_days) | !is.na(EFS_days))

retain_clusters_hgg <- hgg %>%
  dplyr::count(cluster) %>%
  filter(n >= 10) %>%
  pull(cluster)

hgg <- hgg %>%
  filter(cluster %in% retain_clusters_hgg) %>%
  dplyr::mutate(cluster = factor(cluster)) %>%
    dplyr::mutate(SI_group = case_when(
      SBI > summary(SBI)["3rd Qu."] ~ "High SBI",
      SBI < summary(SBI)["1st Qu."] ~ "Low SBI",
      TRUE ~ NA_character_
    )) %>%
  dplyr::mutate(spliceosome_group = case_when(
      spliceosome_gsva_score > summary(spliceosome_gsva_score)["3rd Qu."] ~ "Splice GSVA 4th Q",
      spliceosome_gsva_score > summary(spliceosome_gsva_score)["Median"] ~ "Splice GSVA 3rd Q",
      spliceosome_gsva_score > summary(spliceosome_gsva_score)["1st Qu."] ~ "Splice GSVA 2nd Q",
      TRUE ~ "Splice GSVA 1st Q"
    )) %>%
  dplyr::mutate(SI_group = fct_relevel(SI_group,
                                                 c("High SBI", "Low SBI"))) %>%
  dplyr::mutate(spliceosome_group = fct_relevel(spliceosome_group,
                                                 c("Splice GSVA 1st Q", 
                                                   "Splice GSVA 2nd Q", 
                                                   "Splice GSVA 3rd Q",
                                                   "Splice GSVA 4th Q")))

Generate HGG KM models with spliceosome_group as covariate

# Generate kaplan meier survival models for OS and EFS, and save outputs
hgg_kap_os <- survival_analysis(
  metadata  = hgg %>% dplyr::filter(!is.na(spliceosome_group)),
  ind_var = "spliceosome_group",
  test = "kap.meier",
  metadata_sample_col = "Kids_First_Biospecimen_ID",
  days_col = "OS_days",
  status_col = "OS_status"
)

Testing model: survival::Surv(OS_days, OS_status) ~ spliceosome_group with kap.meier

readr::write_rds(hgg_kap_os,
                 file.path(results_dir, "logrank_hgg_OS_splice_group.RDS"))

hgg_kap_efs <- survival_analysis(
  metadata  = hgg %>% dplyr::filter(!is.na(spliceosome_group)),
  ind_var = "spliceosome_group",
  test = "kap.meier",
  metadata_sample_col = "Kids_First_Biospecimen_ID",
  days_col = "EFS_days",
  status_col = "EFS_status"
)

Testing model: survival::Surv(EFS_days, EFS_status) ~ spliceosome_group with kap.meier

readr::write_rds(hgg_kap_efs,
                 file.path(results_dir, "logrank_hgg_EFS_splice_group.RDS"))

Generate KM plots

km_hgg_os_plot <- plotKM(model = hgg_kap_os,
                    variable = "spliceosome_group",
                    combined = F, 
                    title = "HGG, overall survival",
                    p_pos = "topright")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

ggsave(file.path(plot_dir, "km_hgg_OS_spliceosome_score.pdf"),
       km_hgg_os_plot,
       width = 9, height = 5, units = "in",
       device = "pdf")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Ignoring unknown labels:
• colour : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's colour values.
Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

km_hgg_efs_plot <- plotKM(model = hgg_kap_efs,
                    variable = "spliceosome_group",
                    combined = F, 
                    title = "HGG, event-free survival",
                    p_pos = "topright")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

ggsave(file.path(plot_dir, "km_hgg_EFS_spliceosome_score.pdf"), 
       km_hgg_efs_plot,
       width = 9, height = 5, units = "in",
       device = "pdf")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Ignoring unknown labels:
• colour : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's colour values.
Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Generate coxph models for HGG including covariates mol_sub_group cluster, and SBI, and plot

add_model_hgg_os <- fit_save_model(hgg,
                              terms = "mol_sub_group+age_at_diagnosis_years+SBI",
                               file.path(results_dir, "cox_hgg_OS_additive_terms_subtype_cluster_SBI.RDS"),
                               "multivariate",
                               years_col = "OS_years",
                               status_col = "OS_status")

Warning in coxph.fit(X, Y, istrat, offset, init, control, weights = weights,  :
  Loglik converged before variable  9 ; coefficient may be infinite. 

forest_hgg_os <- plotForest(readRDS(file.path(results_dir, "cox_hgg_OS_additive_terms_subtype_cluster_SBI.RDS")))

`height` was translated to `width`.

Warning: Removed 1 row containing missing values or values outside the scale range (`geom_text()`).

forest_hgg_os

ggsave(file.path(plot_dir, "forest_add_OS_HGG_subtype_cluster_assignment_SBI.pdf"),
       forest_hgg_os,
       width = 9, height = 5, units = "in",
       device = "pdf")

add_model_hgg_efs <- fit_save_model(hgg,
                              terms = "mol_sub_group+age_at_diagnosis_years+SBI",
                               file.path(results_dir, "cox_hgg_EFS_additive_terms_subtype_cluster_SBI.RDS"),
                               "multivariate",
                               years_col = "EFS_years",
                               status_col = "EFS_status")

forest_hgg_efs <- plotForest(readRDS(file.path(results_dir, "cox_hgg_EFS_additive_terms_subtype_cluster_SBI.RDS")))

`height` was translated to `width`.

Warning: Removed 1 row containing missing values or values outside the scale range (`geom_text()`).

ggsave(file.path(plot_dir, "forest_add_EFS_HGG_subtype_cluster_assignment_SBI.pdf"),
       forest_hgg_efs,
       width = 9, height = 5, units = "in",
       device = "pdf")

Repeat analysis replacing SBI with spliceosome_gsva_score

add_model_hgg_os <- fit_save_model(hgg,
                              terms = "mol_sub_group+age_at_diagnosis_years+spliceosome_gsva_score",
                               file.path(results_dir, "cox_hgg_OS_additive_terms_subtype_cluster_spliceosome_score.RDS"),
                               "multivariate",
                               years_col = "OS_years",
                               status_col = "OS_status")

Warning in coxph.fit(X, Y, istrat, offset, init, control, weights = weights,  :
  Loglik converged before variable  9 ; coefficient may be infinite. 

forest_hgg_os <- plotForest(readRDS(file.path(results_dir, "cox_hgg_OS_additive_terms_subtype_cluster_spliceosome_score.RDS")))

`height` was translated to `width`.

Warning: Removed 1 row containing missing values or values outside the scale range (`geom_text()`).

forest_hgg_os

ggsave(file.path(plot_dir, "forest_add_OS_HGG_subtype_cluster_assignment_spliceosome_score.pdf"),
       forest_hgg_os,
       width = 9, height = 5, units = "in",
       device = "pdf")

add_model_hgg_efs <- fit_save_model(hgg,
                              terms = "mol_sub_group+age_at_diagnosis_years+spliceosome_gsva_score",
                               file.path(results_dir, "cox_hgg_EFS_additive_terms_subtype_cluster_spliceosome_score.RDS"),
                               "multivariate",
                               years_col = "EFS_years",
                               status_col = "EFS_status")

forest_hgg_efs <- plotForest(readRDS(file.path(results_dir, "cox_hgg_EFS_additive_terms_subtype_cluster_spliceosome_score.RDS")))

`height` was translated to `width`.

Warning: Removed 1 row containing missing values or values outside the scale range (`geom_text()`).

ggsave(file.path(plot_dir, "forest_add_EFS_HGG_subtype_cluster_assignment_spliceosome_score.pdf"),
       forest_hgg_efs,
       width = 9, height = 5, units = "in",
       device = "pdf")

Filter for cluster 7

cluster7_df <- metadata %>%
  dplyr::filter(cluster == "Cluster 7",
                !is.na(EFS_days)) %>%
  dplyr::mutate(SI_group = case_when(
      SBI > summary(SBI)["3rd Qu."] ~ "High SBI",
      SBI < summary(SBI)["1st Qu."] ~ "Low SBI",
      TRUE ~ NA_character_
    )) %>%
  dplyr::mutate(spliceosome_group = case_when(
      spliceosome_gsva_score > summary(spliceosome_gsva_score)["3rd Qu."] ~ "Splice GSVA 4th Q",
      spliceosome_gsva_score > summary(spliceosome_gsva_score)["Median"] ~ "Splice GSVA 3rd Q",
      spliceosome_gsva_score > summary(spliceosome_gsva_score)["1st Qu."] ~ "Splice GSVA 2nd Q",
      TRUE ~ "Splice GSVA 1st Q"
    )) %>%
  dplyr::mutate(SI_group = fct_relevel(SI_group,
                                                 c("High SBI", "Low SBI"))) %>%
  dplyr::mutate(spliceosome_group = fct_relevel(spliceosome_group,
                                                 c("Splice GSVA 1st Q", 
                                                   "Splice GSVA 2nd Q", 
                                                   "Splice GSVA 3rd Q",
                                                   "Splice GSVA 4th Q")))

Generate KM models with SI_group as covariate

# Generate kaplan meier survival models for OS and EFS, and save outputs
c7_si_kap_os <- survival_analysis(
  metadata  = cluster7_df %>% dplyr::filter(!is.na(SI_group)),
  ind_var = "SI_group",
  test = "kap.meier",
  metadata_sample_col = "Kids_First_Biospecimen_ID",
  days_col = "OS_days",
  status_col = "OS_status"
)

Testing model: survival::Surv(OS_days, OS_status) ~ SI_group with kap.meier

readr::write_rds(c7_si_kap_os,
                 file.path(results_dir, "logrank_cluster7_OS_SBI.RDS"))

c7_si_kap_efs <- survival_analysis(
  metadata  = cluster7_df %>% dplyr::filter(!is.na(SI_group)),
  ind_var = "SI_group",
  test = "kap.meier",
  metadata_sample_col = "Kids_First_Biospecimen_ID",
  days_col = "EFS_days",
  status_col = "EFS_status"
)

Testing model: survival::Surv(EFS_days, EFS_status) ~ SI_group with kap.meier

readr::write_rds(c7_si_kap_efs,
                 file.path(results_dir, "logrank_cluster7_EFS_SBI.RDS"))

Generate Cluster 7 KM SI_group plots

km_c7_si_os_plot <- plotKM(model = c7_si_kap_os,
                    variable = "SI_group",
                    combined = F, 
                    title = "Cluster 7, overall survival",
                    p_pos = "topright")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

ggsave(file.path(plot_dir, "km_cluster7_OS_sbi_group.pdf"),
       km_c7_si_os_plot,
       width = 8, height = 5, units = "in",
       device = "pdf")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Ignoring unknown labels:
• colour : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's colour values.
Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

km_c7_si_efs_plot <- plotKM(model = c7_si_kap_efs,
                    variable = "SI_group",
                    combined = F, 
                    title = "Cluster 7, event-free survival",
                    p_pos = "topright")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

ggsave(file.path(plot_dir, "km_cluster7_EFS_sbi_group.pdf"), 
       km_c7_si_efs_plot,
       width = 8, height = 5, units = "in",
       device = "pdf")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Ignoring unknown labels:
• colour : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's colour values.
Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Generate KM models with spliceosome_group as covariate

# Generate kaplan meier survival models for OS and EFS, and save outputs
c7_splice_kap_os <- survival_analysis(
  metadata  = cluster7_df %>% 
    dplyr::filter(spliceosome_group %in% c("Splice GSVA 4th Q", "Splice GSVA 1st Q")) %>%
    dplyr::mutate(spliceosome_group = factor(spliceosome_group,
                                                  levels = c("Splice GSVA 1st Q", "Splice GSVA 4th Q"))),
  ind_var = "spliceosome_group",
  test = "kap.meier",
  metadata_sample_col = "Kids_First_Biospecimen_ID",
  days_col = "OS_days",
  status_col = "OS_status"
)

Testing model: survival::Surv(OS_days, OS_status) ~ spliceosome_group with kap.meier

readr::write_rds(c7_splice_kap_os,
                 file.path(results_dir, "logrank_cluster7_OS_splice_group.RDS"))

c7_splice_kap_efs <- survival_analysis(
  metadata  = cluster7_df %>% 
    dplyr::filter(spliceosome_group %in% c("Splice GSVA 4th Q", "Splice GSVA 1st Q")) %>%
    dplyr::mutate(spliceosome_group = factor(spliceosome_group,
                                                  levels = c("Splice GSVA 1st Q", "Splice GSVA 4th Q"))),
  ind_var = "spliceosome_group",
  test = "kap.meier",
  metadata_sample_col = "Kids_First_Biospecimen_ID",
  days_col = "EFS_days",
  status_col = "EFS_status"
)

Testing model: survival::Surv(EFS_days, EFS_status) ~ spliceosome_group with kap.meier

readr::write_rds(c7_splice_kap_efs,
                 file.path(results_dir, "logrank_cluster7_EFS_splice_group.RDS"))

Generate Cluster 7 KM spliceosome_group plots

km_c7_splice_os_plot <- plotKM(model = c7_splice_kap_os,
                    variable = "spliceosome_group",
                    combined = F, 
                    title = "Cluster 7, overall survival",
                    p_pos = "topright")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

ggsave(file.path(plot_dir, "km_cluster7_OS_splice_group.pdf"),
       km_c7_splice_os_plot,
       width = 9, height = 5, units = "in",
       device = "pdf")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Ignoring unknown labels:
• colour : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's colour values.
Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

km_c7_splice_efs_plot <- plotKM(model = c7_splice_kap_efs,
                    variable = "spliceosome_group",
                    combined = F, 
                    title = "Cluster 7, event-free survival",
                    p_pos = "topright")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

ggsave(file.path(plot_dir, "km_cluster7_EFS_splice_group.pdf"), 
       km_c7_splice_efs_plot,
       width = 9, height = 5, units = "in",
       device = "pdf")

Ignoring unknown labels:
• fill : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Ignoring unknown labels:
• colour : ""

Warning: No shared levels found between `names(values)` of the manual scale and the data's colour values.
Warning: No shared levels found between `names(values)` of the manual scale and the data's fill values.

Assess EFS and OS by SBI or spliceosome GSVA score in multivariate models and generate forest plots

add_model_c7_efs <- fit_save_model(cluster7_df %>% 
                                      dplyr::filter(extent_of_tumor_resection != "Unavailable",
                                                    spliceosome_group %in% c("Splice GSVA 4th Q", "Splice GSVA 1st Q")) %>%
                                      dplyr::mutate(plot_group = fct_relevel(plot_group, "Low-grade glioma", after = 0)),
                              terms = "extent_of_tumor_resection+age_at_diagnosis_years+plot_group+spliceosome_group",
                               file.path(results_dir, "cox_hgg_EFS_additive_terms_subtype_cluster_spliceosome_score.RDS"),
                               "multivariate",
                               years_col = "EFS_years",
                               status_col = "EFS_status")

Warning in coxph.fit(X, Y, istrat, offset, init, control, weights = weights,  :
  Loglik converged before variable  6 ; coefficient may be infinite. 

forest_c7_spliceosome_efs <- plotForest(readRDS(file.path(results_dir, "cox_hgg_EFS_additive_terms_subtype_cluster_spliceosome_score.RDS")))

`height` was translated to `width`.

Warning: Removed 3 rows containing missing values or values outside the scale range (`geom_text()`).

ggsave(file.path(plot_dir, "forest_add_EFS_cluster7_histology_resection_spliceosome_group.pdf"),
       forest_c7_spliceosome_efs,
       width = 9, height = 4, units = "in",
       device = "pdf")

add_model_c7_os <- fit_save_model(cluster7_df %>% 
                                    dplyr::filter(!extent_of_tumor_resection %in% c("Not Reported", "Unavailable")) %>%
                                    dplyr::mutate(plot_group = fct_relevel(plot_group, "Low-grade glioma", after = 0)),
                              terms = "extent_of_tumor_resection+age_at_diagnosis_years+plot_group+SBI",
                               file.path(results_dir, "cox_hgg_OS_additive_terms_subtype_cluster_si_group.RDS"),
                               "multivariate",
                               years_col = "OS_years",
                               status_col = "OS_status")

Warning in coxph.fit(X, Y, istrat, offset, init, control, weights = weights,  :
  Loglik converged before variable  4,5,8,9 ; coefficient may be infinite. 

forest_c7_si_os <- plotForest(readRDS(file.path(results_dir, "cox_hgg_OS_additive_terms_subtype_cluster_si_group.RDS")))

Warning in scale_x_log10(labels = function(x) format(x, scientific = FALSE)) :
  log-10 transformation introduced infinite values.

`height` was translated to `width`.

Warning: Removed 2 rows containing missing values or values outside the scale range (`geom_text()`).

ggsave(file.path(plot_dir, "forest_add_OS_cluster7_histology_resection_si.pdf"),
       forest_c7_si_os,
       width = 9, height = 4, units = "in",
       device = "pdf")

Print session info

sessionInfo()

R version 4.4.0 (2024-04-24)
Platform: x86_64-pc-linux-gnu
Running under: Ubuntu 22.04.4 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.20.so;  LAPACK version 3.10.0

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C               LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8     LC_MONETARY=en_US.UTF-8   
 [6] LC_MESSAGES=en_US.UTF-8    LC_PAPER=en_US.UTF-8       LC_NAME=C                  LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

time zone: Etc/UTC
tzcode source: system (glibc)

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] gtools_3.9.5    survminer_0.4.9 patchwork_1.2.0 ggpubr_0.6.0    survival_3.7-0  lubridate_1.9.4 forcats_1.0.1   stringr_1.6.0  
 [9] dplyr_1.1.4     purrr_1.2.0     readr_2.1.6     tidyr_1.3.1     tibble_3.3.0    ggplot2_4.0.1   tidyverse_2.0.0

loaded via a namespace (and not attached):
 [1] gtable_0.3.6       xfun_0.54          bslib_0.9.0        rstatix_0.7.2      lattice_0.22-7     tzdb_0.5.0         vctrs_0.6.5       
 [8] tools_4.4.0        generics_0.1.4     parallel_4.4.0     pkgconfig_2.0.3    Matrix_1.7-4       data.table_1.17.8  RColorBrewer_1.1-3
[15] S7_0.2.1           lifecycle_1.0.4    compiler_4.4.0     farver_2.1.2       textshaping_1.0.4  carData_3.0-5      colorblindr_0.1.0 
[22] htmltools_0.5.8.1  sass_0.4.10        yaml_2.3.10        crayon_1.5.3       pillar_1.11.1      car_3.1-2          jquerylib_0.1.4   
[29] cachem_1.1.0       abind_1.4-5        km.ci_0.5-6        commonmark_2.0.0   tidyselect_1.2.1   digest_0.6.39      stringi_1.8.7     
[36] labeling_0.4.3     splines_4.4.0      cowplot_1.1.3      rprojroot_2.1.1    fastmap_1.2.0      grid_4.4.0         colorspace_2.1-2  
[43] cli_3.6.5          magrittr_2.0.4     broom_1.0.10       withr_3.0.2        scales_1.4.0       backports_1.5.0    bit64_4.6.0-1     
[50] timechange_0.3.0   rmarkdown_2.30     ggtext_0.1.2       bit_4.6.0          gridExtra_2.3      ggsignif_0.6.4     ragg_1.5.0        
[57] zoo_1.8-12         hms_1.1.4          evaluate_1.0.5     knitr_1.50         KMsurv_0.1-5       markdown_1.13      survMisc_0.5.6    
[64] rlang_1.1.6        Rcpp_1.1.0         gridtext_0.1.5     xtable_1.8-4       glue_1.8.0         xml2_1.5.0         vroom_1.6.6       
[71] rstudioapi_0.17.1  jsonlite_2.0.0     R6_2.6.1           systemfonts_1.3.1